A multiobjective evolutionary algorithm based on decomposition for hybrid flowshop green scheduling problem. (October 2019)

• Record Type: Journal Article
• Authors: Zhang, Biao; Pan, Quan-ke; Gao, Liang; Li, Xin-yu; Meng, Lei-lei; Peng, Kun-kun
• Is Part Of: Computers & industrial engineering. Volume 136(2019)
• Page Start: 325
• Page End: 344
• Publication Date: 2019-10
• Subjects: Green scheduling -- Hybrid flowshop -- Artificial bee colony algorithm -- Decomposition -- Multiobjective optimization; Engineering -- Data processing -- Periodicals; Industrial engineering -- Periodicals; 620.00285
• DOI: 10.1016/j.cie.2019.07.036
• Languages: English
• ISSNs: 0360-8352

Abstract

Highlights: A hybrid flowshop green scheduling problem with various processing speeds is modeled. A multiobjective discrete artificial bee colony algorithm based on decomposition is proposed. Several definitions are proposed to implement objective normalization. A dynamic sub-problem neighborhood and a solution exchange strategy are proposed. An energy-saving procedure is developed. Abstract: Energy saving has attracted growing attention due to the advent of sustainable manufacturing. By this motivation, this paper studies a hybrid flowshop green scheduling problem (HFGSP) with variable machine processing speeds. A multi-objective optimization model with the objectives of minimizing the makespan and total energy consumption is developed. To solve this complex problem, a multiobjective discrete artificial bee colony algorithm (MDABC) based on decomposition is suggested. In VND-based employed bee phase, the variable neighborhood descent (VND) with five designed neighborhood is employed to each subproblem to realize their self-evolution. In the collaborative onlooker bee phase, the promising subproblems selected by the order preference technique according to their similarity to an ideal solution (TOPSIS) is evolved by collaborating with the other neighboring subproblems. Particularly, a dynamic neighborhood strategy is developed to define the neighborhood relationship to retain the population diversity. In the solution exchange-based scout bee phase, a solution exchange strategy is developed to enhance the algorithm efficiency and enable the solutions to be exploited in different directions. Moreover, according to the problem-specific characteristics, encoding and decoding methodologies are developed to represent the solution space, and several definitions are proposed to implement objective normalization, and an energy saving procedure is designed to reduce the energy consumption. Through comprehensive computational comparisons and statistical analysis, the developed strategies and MDABC shows highly effective performance.